Multi-Tape Two-Level Morphology: A Case Study in Semitic Non-Linear Morphology

What it is about

Semitic stems are built by interleaving a consonantal root with a vowel melody inside a CV pattern, and standard two-level morphology (Koskenniemi 1983) cannot describe that directly. Kiraz keeps the two-level idea of one surface tape but lets the lexical side have several tapes: a pattern tape, a root tape, a vocalism tape, and a primary tape for prefixes, suffixes and particles. A lexical expression becomes an n-tuple regular expression, and an ellipsis marks the optional omission of tuples from the left context, which models spreading. The formalism is applied to the Arabic verb (forms I–X, active and passive) and compiled into an auxiliary multi-tape automaton. The part of the paper we read has no evaluation figures; its claim is about description and coverage.

Ideas

I1. Binyanim as tapes: a multi-tape rule set for Hebrew with a root tape (consonants), a pattern tape (the CV template of Qal, Niphal, Piel, Pual, Hiphil, Hophal, Hithpael) and a vocalism tape, projected onto the surface. The segmenter in tools/hbo splits affixes only, so the root tape would be a new stage between lex and gparse.
I2. Ellipsis as defective spelling: allow the optional omission of weak radicals (the consonants aleph, waw and yod) in the left context. This is the same device needed for weak verbs, where a radical is written once or not at all.
I3. Use the paper's Arabic verb table (forms I–X, active and passive) as a small gold set of root and pattern pairs. Rebuild the same stems with the Hebrew binyanim as a test of the multi-tape generator, not as a release item.

For Mova

I1 goes into The verb under the binyanim; I2 goes into The weak verbs; I3 is a test for tools/hbo.

Cite

The paper itself, not this card. BibTeX from the library's record of the work.

@inproceedings{kiraz1994multi,
  title = {{Multi-Tape Two-Level Morphology: A Case Study in Semitic Non-Linear Morphology}},
  author = {Kiraz, George Anton},
  booktitle = {COLING 1994 Volume 1: The 15th International Conference on Computational Linguistics},
  year = {1994},
  url = {https://aclanthology.org/C94-1029/}
}

References

2 references, 0 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

I(. Beesley. 'Computer Analysis of Aral)ic Morphology.' 11. Comrie and M. l"id (eds.) l'erspectives on Arabic Linguistics IlL []~ird ~l.l~ld ]~][]~()[l 1992] S. ]~hd and T. l']llison. One Level Phonvlogy. Edinburgh research Papers in Cognitive Science, No. EU('CS/RP~51 (updated version 1993). find
A. Black, G. Ritchie, S. Puhnan, G. ]hlssel. 'Formalisms fbr Morphographernlc Description.' 1,2ACL-3. [(:oidsmith 1976] J. (',oMsmith. Autosegmental Iqwnology, l)oct()r;t] dissertaLion, MIT, Published later ,'us Autosegmental and Metrical l'honology (Ox[ord 1(.)90.) [llarris 1941) Z. llm'ris. 'Linguistic Structmre of Hebrew.' ,liner-nag of the American Oriental 5'oclety: 61. find

This page in Markdown → https://md.mova.elamur.ai/free/en/docs/papers/aclbulk-1994-kiraz-multi-tape-two-level-morphology/